Introduction

The intention with this power amp is to get cranked amp sound at very low levels, without using a speaker attenuator. Most of us tube amp guitar players got other people in the house or in the neighborhood, and disappointingly have to adjust the volume thereafter. 

There have been discussions about this. Some say you'll never get the right feeling of a cranked amp at low volume, because that cranked tone comes from both an overdriven power amp and a distorted speaker, and even a saturated output transformer. And they may have right about that. You get an overdriven power amp but not that distorted speaker as when you push 50 watt into it.

As a matter of fact it's impossible to get it all at low volume. The question then is, should you have no cranked sound at low volume at all, or would you like something close up to it?

I've compared this "power reduced" amp with lots of different preamp overdrive sounds, which is the alternative. This is far more bluesy, Angus Young'ish than you'll ever get out of any preamp overdrive.

I strongly recommed to download both the conventional EL34 push-pull schematic and the power reduction one, for the easiest way to see the differences.

The whole point with a power control circuit is to lower the power amp’s headroom, so overdrive occurs at a lower output level. This means the power control does nothing to the output volume, it rather controls where or when the amp should crank.

Adjustable supply voltage

The headroom can be reduced by lowering the B+ voltage in to the power tubes (anode and screen). In this amp this is done by a power MOSFET transistor (T101). A 470k linear pot (P102A) feeds a variable voltage to the MOSFET’s gate. The drain makes a voltage that follows the variable gate voltage. The drain voltage can then be adusted between 400V down to 56V. The lowest voltage is set by R122.

NOTE! The MOSFET must be mounted on an aluminium cooler or similar.

The variable voltage is fed through the output transformer’s primary side, feeding the anodes on the power tubes. To feed the screen grids an additional choke coil and a filter cap is used to prevent hum.

There is actually two chokes in this circuit. One for the variable voltage (screen grids) and one for the fixed 400V feeding the phase inverter, and the preamp that is wished.

Bias compensation

When the voltage to the power tubes is reduced a bias problem occurs. The bias current decreases when the variable +B voltage is changed. This means the fixed bias voltage has to be compensated to keep the bias current correct at every voltage level of the +B.

The fixed bias voltage has to be reduced proportionally to the anode voltage. This is solved in a little tricky way. A 100k linear pot P102B is placed in the bias circuit and it is physically ganged to the 470k pot (R102A) that’s feeding the MOSFET.

This means you have to make a dual ganged pot yourself. Take a 470k dual ganged linear pot and replace one the of carbon lanes (at the back of the pot) with a physically equal 100k linear lane.

At the lowest anode voltage the power tubes are completely cathode biased. In this first version of the amp two 1ohm resistors (R111 and R112) take care of this biasing. The 1ohm resistors was originally placed in the circuit for measuring the bias current in a simple way (in an ordinary push pull power amp), and had really nothing to do with the variable power circuit. As a matter of fact the resistors should be changed to a lower value because the bias current at 56V anode voltage is only 18mA, and the amp is sounding a bit "cold" at the lowest output power. Lack of time is the reason I don’t have tried this yet.

SAG-switch

The intention with this is to simulate a tube rectifier’s inner resistance that SS rectifiers don’t have. This makes more sag at high output power, when there is pulled a lot of current from the power supply. This resistor has hardly some effect at low power output.